Abstract: For the current inaccuracy in passenger subdivision of airlines, this article proposes a passenger value evaluation model to improve the accuracy of customer segmentation based on traditional models. To subdivide passengers depending on their behavioral preference, this model is constructed through Analytic Hierarchy Process（AHP）, where how to build consistent comparison matrices is of key importance in determining the weight of various parameters. An algorithm is promoted in this article to enhance the consistency of comparison matrices to achieve simpleness in consistency checking and efficiency in results. By adopting the improved AHP to obtain weights for parameters and segmenting customers via ticket information, airlines are provided valuable references in offering personalized services to passengers with different preferences and developing effective marketing strategy.

Key words: passenger subdivision, civil aviation passenger, Analytic Hierarchy Process（AHP）, clustering analysis

摘要： 针对当前航空公司旅客细分工作不够精准的问题，在传统旅客划分模型的基础上，提出了一种旅客价值评价模型来提高旅客细分的准确度，该模型根据旅客的行为偏好对旅客进行细分，构建模型采用的方法为层次分析法，其中在使用层次分析法确定各参数权重时的关键在于如何构造完全一致的判断矩阵，提出了一种改进判断矩阵一致性的算法，使得一致性调整过程简单且结果有效。采用改进后的层次分析法确定旅客模型各参数的权重并通过旅客购票信息对旅客进行分类，为航空公司针对有着不同行为偏好的旅客制定相应的个性化服务以及营销策略提供了良好的参考依据。

关键词: 客户细分, 航空旅客, 层次分析法, 聚类分析